基于图像识别的梭子蟹抓取研究  

Research on the Capture of Portunus trituberculatus Based on Image Recognition

作　　者：张传龙 徐鑫 胡海刚[1] Zhang Chuanlong;Xu Xin;Hu Haigang(Faculty of Marine and Transportation,Ningbo University,Ningbo 315832,Zhejiang,China)

机构地区：[1]宁波大学海运学院,浙江宁波315832

出　　处：《农业技术与装备》2025年第1期20-24,共5页Agricultural Technology & Equipment

基　　金：浙江省科学技术厅基础公益项目(LGN22C190001)。

摘　　要：面对梭子蟹养殖行业中人力收捕作业效率低下和成本高昂的挑战,基于机器视觉和机械自动化技术,进行了图像识别驱动的梭子蟹抓取研究,通过考虑相机标定和机械臂坐标变换的关键因素,提出了一种基于深度学习的图像识别方法与仿射变换相结合的模型,用于将相机视野中的梭子蟹位置映射至机械臂的操作空间,以方便抓取。研究成果不仅降低了人工成本,而且提升了抓取效率,为梭子蟹自动化养殖技术的发展奠定了基础。In the face of the challenges of low efficiency and high cost of human harvesting in the culture industry of Portunus tritu‐berculatus,a research on the capture of Portunus trituberculatus driven by image recognition was carried out based on machine vi‐sion and mechanical automation technologies.By considering the key factors of camera calibration and robotic arm coordinate trans‐formation,a model combining image recognition method based on deep learning and affine transformation was proposed.It was used to map the position of Portunus trituberculatus in the camera field of view to the operating space of the robot arm for easy grasping.The research results not only reduce the labor cost,but also improve the grasping efficiency,and lay a foundation for the develop‐ment of automatic culture technology of Portunus trituberculatus.

关 键 词：梭子蟹 自动化养殖 梭子蟹抓取 图像识别 机械臂 

分 类 号：TH-39[机械工程] TP391.4[自动化与计算机技术—计算机应用技术]